Day 17 - Almas Fathin Irbah - Machine Learning: Unsupervised Learning with R
library(tidyverse) # data manipulation and visualization## -- Attaching packages --------------------------------------- tidyverse 1.3.1 --## v ggplot2 3.3.3 v purrr 0.3.4
## v tibble 3.1.2 v dplyr 1.0.6
## v tidyr 1.1.3 v stringr 1.4.0
## v readr 1.4.0 v forcats 0.5.1## -- Conflicts ------------------------------------------ tidyverse_conflicts() --
## x dplyr::filter() masks stats::filter()
## x dplyr::lag() masks stats::lag()library(readr)
#install.packages("gridExtra")
library(gridExtra) # plot arrangement##
## Attaching package: 'gridExtra'## The following object is masked from 'package:dplyr':
##
## combinecustomer_movie_rating <- read.csv("customer_movie_rating.csv")
print(customer_movie_rating)## Horror Romcom Action Comedy Fantasy
## 1 72.5 29.9 68.6 40.7 57.9
## 2 82.2 45.3 76.5 17.4 67.7
## 3 70.0 44.0 65.1 53.7 37.8
## 4 99.1 21.0 77.9 25.4 40.3
## 5 84.0 0.0 68.1 49.8 40.0
## 6 70.2 55.0 97.2 48.1 40.5
## 7 85.8 33.0 84.6 31.0 40.6
## 8 88.9 31.5 86.9 36.8 41.5
## 9 86.9 24.8 80.6 17.7 62.3
## 10 76.3 31.1 96.2 33.9 36.6
## 11 98.1 23.0 68.4 42.5 40.3
## 12 84.7 30.0 70.5 31.2 52.8
## 13 72.5 20.2 93.2 24.6 58.5
## 14 53.4 8.6 68.2 22.1 35.7
## 15 93.5 36.9 73.5 29.6 38.9
## 16 79.5 43.2 71.3 42.8 51.3
## 17 79.8 21.3 72.2 24.2 57.2
## 18 91.3 10.0 72.7 28.5 42.0
## 19 89.9 32.7 81.9 14.5 27.8
## 20 87.1 24.5 73.9 35.9 65.5
## 21 91.0 4.2 69.9 45.7 62.2
## 22 89.4 25.0 92.1 48.0 36.5
## 23 80.9 17.4 73.4 39.8 44.2
## 24 56.1 20.9 73.8 7.6 23.9
## 25 87.4 11.1 74.8 35.8 51.8
## 26 79.3 46.6 84.6 35.5 64.3
## 27 78.1 21.0 75.1 25.8 25.4
## 28 62.4 5.7 75.5 32.0 35.6
## 29 74.3 27.4 67.8 19.6 37.3
## 30 85.0 28.2 72.1 38.2 36.8
## 31 96.3 13.2 76.7 26.1 20.6
## 32 78.8 0.0 68.9 11.2 51.0
## 33 84.7 17.3 82.4 25.6 26.6
## 34 79.4 31.8 57.8 46.4 44.7
## 35 63.5 24.3 79.7 26.0 52.1
## 36 75.0 23.8 57.6 38.8 42.6
## 37 75.3 31.7 72.4 41.4 55.7
## 38 79.3 10.8 69.7 30.1 44.7
## 39 93.2 38.2 68.2 25.8 37.2
## 40 89.2 24.9 75.3 23.6 52.8
## 41 78.0 33.5 53.0 38.9 39.8
## 42 77.0 37.4 90.1 17.2 66.3
## 43 88.4 27.7 56.0 33.0 44.8
## 44 86.7 14.5 70.4 26.5 55.2
## 45 71.7 39.0 62.6 2.8 47.5
## 46 71.5 1.0 67.0 13.1 8.9
## 47 84.4 18.5 100.0 41.0 28.6
## 48 89.2 21.9 76.2 27.7 39.9
## 49 78.7 23.0 60.6 39.6 47.8
## 50 90.6 37.2 56.3 52.6 16.9
## 51 84.8 26.6 81.4 47.7 56.5
## 52 72.7 29.9 75.8 38.1 37.7
## 53 84.1 24.2 72.2 34.6 36.0
## 54 66.4 22.0 64.8 27.7 26.3
## 55 97.2 33.3 58.2 48.9 27.6
## 56 100.0 38.8 63.1 37.2 45.7
## 57 75.6 0.0 88.0 15.9 51.1
## 58 67.5 31.9 68.5 28.1 19.8
## 59 86.8 29.5 59.4 7.0 32.9
## 60 78.4 19.9 98.4 27.7 51.4
## 61 100.0 36.4 81.1 0.0 39.6
## 62 79.5 20.3 73.1 45.8 71.0
## 63 88.3 21.6 88.7 22.4 59.9
## 64 80.3 35.3 86.6 24.8 52.1
## 65 71.1 45.6 68.6 28.0 45.1
## 66 82.3 28.2 100.0 37.3 48.4
## 67 58.3 19.9 72.9 38.1 36.5
## 68 97.6 10.7 58.9 36.8 52.5
## 69 81.8 21.0 74.3 23.1 62.8
## 70 100.0 13.7 78.5 13.6 58.0
## 71 85.7 21.9 100.0 25.3 35.2
## 72 71.5 29.7 77.3 33.3 25.6
## 73 87.3 14.8 81.5 20.1 43.7
## 74 68.8 56.8 75.1 29.2 50.3
## 75 65.0 26.9 72.0 16.0 61.2
## 76 83.5 38.6 75.6 29.9 29.2
## 77 74.7 0.0 85.5 31.5 49.4
## 78 80.0 33.9 100.0 28.2 47.8
## 79 80.9 9.2 88.3 28.0 59.3
## 80 72.9 36.0 90.5 51.2 44.7
## 81 73.2 29.8 61.2 39.2 40.7
## 82 78.4 20.1 87.8 43.3 31.2
## 83 94.1 40.9 78.6 18.9 38.8
## 84 61.7 16.6 58.4 32.0 40.7
## 85 87.1 18.0 82.3 43.9 73.2
## 86 84.0 13.0 74.1 29.3 74.4
## 87 92.8 17.0 93.6 4.4 43.0
## 88 76.3 36.3 66.8 34.1 32.5
## 89 84.4 30.2 70.8 7.1 21.3
## 90 83.2 37.1 64.9 20.3 51.2
## 91 73.5 20.3 73.9 45.9 31.9
## 92 94.5 29.5 80.8 37.4 72.4
## 93 93.9 27.9 67.2 43.1 34.4
## 94 88.4 7.9 86.0 33.7 46.3
## 95 99.0 46.3 61.5 28.7 90.7
## 96 86.7 26.6 63.4 18.9 31.7
## 97 64.7 34.2 93.3 49.1 48.7
## 98 73.1 36.5 63.8 30.5 31.7
## 99 65.3 24.4 80.9 21.4 49.2
## 100 74.3 21.3 71.4 40.4 34.5
## 101 25.9 25.8 38.1 93.8 65.9
## 102 21.4 50.9 60.2 90.6 88.8
## 103 10.8 63.6 59.7 72.1 64.9
## 104 25.1 47.1 44.4 60.3 79.7
## 105 36.9 54.6 60.2 71.1 85.7
## 106 44.1 35.9 61.8 46.1 59.0
## 107 8.5 65.2 56.9 71.2 79.8
## 108 22.0 32.8 74.3 94.9 73.7
## 109 38.9 51.1 26.5 76.6 83.1
## 110 11.6 51.9 36.0 88.2 65.5
## 111 0.0 55.7 33.5 80.9 74.0
## 112 13.8 45.1 52.0 65.7 91.6
## 113 13.4 77.0 69.0 95.9 77.0
## 114 25.6 37.8 70.1 74.1 84.9
## 115 20.2 58.0 55.9 63.3 72.3
## 116 27.8 19.7 60.5 75.8 62.6
## 117 19.9 55.0 48.3 56.8 74.9
## 118 29.5 61.1 55.6 85.4 60.3
## 119 20.6 42.0 54.5 81.0 43.8
## 120 39.3 63.5 40.9 70.7 89.0
## 121 16.2 59.0 46.5 69.1 62.0
## 122 28.5 66.7 48.4 86.2 53.1
## 123 14.4 44.9 66.7 62.3 86.9
## 124 27.5 43.4 37.6 63.2 74.9
## 125 24.4 33.7 24.6 80.1 67.8
## 126 4.8 51.1 59.2 69.6 72.9
## 127 23.3 38.9 40.2 86.1 69.9
## 128 39.2 63.3 44.8 72.6 87.0
## 129 33.2 55.9 60.9 89.3 83.7
## 130 26.7 81.3 40.8 80.9 54.3
## 131 10.7 41.1 45.9 48.1 79.2
## 132 39.0 69.2 68.0 59.0 83.7
## 133 26.0 48.7 56.3 90.4 83.0
## 134 19.6 37.5 56.5 83.3 45.9
## 135 44.7 61.9 48.4 63.4 72.2
## 136 15.8 37.8 36.4 58.9 98.8
## 137 28.6 42.3 32.0 87.4 84.6
## 138 40.4 46.2 47.3 65.3 54.5
## 139 32.2 57.5 74.0 96.6 55.0
## 140 21.3 43.0 52.7 96.3 80.9
## 141 20.0 67.9 52.0 57.5 72.9
## 142 29.3 40.6 54.0 64.9 86.5
## 143 31.2 57.6 45.4 60.0 78.5
## 144 25.2 56.7 33.2 83.0 76.0
## 145 40.8 42.2 82.1 59.5 77.2
## 146 24.2 49.9 44.9 50.6 77.0
## 147 16.6 47.1 46.4 99.3 59.8
## 148 31.4 66.5 28.5 87.1 100.0
## 149 0.0 73.7 47.0 84.8 58.1
## 150 29.7 42.5 47.0 67.0 74.8
## 151 31.0 66.2 46.9 74.9 68.5
## 152 22.3 54.1 28.6 82.4 96.6
## 153 11.6 31.4 71.4 59.6 87.1
## 154 20.3 45.9 71.2 73.5 68.2
## 155 43.6 60.5 58.3 77.1 77.5
## 156 16.1 56.7 36.8 95.3 89.0
## 157 0.0 25.7 58.6 82.7 100.0
## 158 34.7 62.0 47.0 90.4 78.6
## 159 19.0 62.9 46.2 76.7 62.5
## 160 18.9 51.3 66.4 61.6 63.2
## 161 10.4 46.5 35.3 70.9 99.1
## 162 24.7 78.7 43.9 55.0 50.2
## 163 33.4 53.9 41.2 86.1 100.0
## 164 18.0 54.8 50.2 92.0 71.9
## 165 17.7 41.5 31.1 74.2 69.5
## 166 38.2 38.9 41.6 63.2 76.9
## 167 22.0 36.7 58.6 88.0 86.2
## 168 23.1 49.9 55.6 76.7 78.6
## 169 17.5 71.3 38.3 70.4 51.5
## 170 35.8 76.0 56.7 88.5 79.2
## 171 65.4 46.4 80.5 45.9 23.1
## 172 67.9 33.5 85.5 80.7 20.1
## 173 47.6 20.9 63.5 67.4 27.1
## 174 31.5 24.9 66.5 60.9 22.9
## 175 56.1 23.5 39.0 80.2 25.1
## 176 48.7 0.0 63.5 76.3 33.9
## 177 50.8 45.4 79.6 80.3 16.0
## 178 48.6 37.8 81.1 61.5 39.1
## 179 55.2 15.8 99.3 74.3 30.0
## 180 56.3 29.6 75.8 68.5 29.5
## 181 69.6 27.9 80.6 58.8 5.6
## 182 71.8 11.4 75.9 86.1 26.4
## 183 50.5 42.9 69.2 66.9 23.3
## 184 56.3 22.0 84.0 64.3 40.2
## 185 64.3 27.2 80.6 71.3 46.7
## 186 76.8 29.9 76.6 69.7 44.8
## 187 59.3 13.1 65.2 56.5 16.6
## 188 39.6 12.0 74.5 64.4 20.4
## 189 62.8 24.4 84.4 96.3 28.9
## 190 58.6 31.9 83.1 64.1 17.3
## 191 81.3 25.9 69.2 61.4 10.7
## 192 64.1 33.5 67.0 73.8 39.5
## 193 52.5 29.0 81.2 69.8 30.8
## 194 54.7 31.5 87.6 58.5 37.4
## 195 53.9 8.2 76.5 52.0 15.6
## 196 62.2 33.1 71.2 75.2 25.9
## 197 62.1 15.5 64.4 66.4 15.9
## 198 71.0 19.4 70.0 70.8 20.0
## 199 63.8 23.7 57.2 61.5 32.8
## 200 55.6 5.2 66.6 45.2 20.2
## 201 48.7 23.8 88.6 65.8 18.1
## 202 67.6 19.7 88.3 93.7 14.3
## 203 59.3 16.4 64.6 47.5 12.8
## 204 62.2 18.3 77.5 53.2 19.6
## 205 73.2 39.9 75.8 49.0 50.4
## 206 81.0 9.9 55.0 60.6 22.1
## 207 48.6 22.4 84.7 81.6 16.6
## 208 79.7 0.0 52.1 63.2 13.1
## 209 49.6 16.9 60.0 61.9 28.2
## 210 63.2 19.0 87.0 49.5 25.3
## 211 62.7 43.5 68.5 65.8 33.0
## 212 56.7 13.5 72.4 77.4 26.6
## 213 76.7 14.5 55.5 92.1 40.4
## 214 52.1 8.2 57.6 80.8 38.6
## 215 67.9 27.2 79.2 54.6 42.7
## 216 59.8 38.8 72.9 59.0 32.9
## 217 48.8 10.6 67.9 72.3 43.4
## 218 74.6 19.9 59.0 64.9 29.1
## 219 34.9 41.4 61.8 68.3 41.2
## 220 53.7 16.8 99.4 49.3 28.1
## 221 41.5 33.2 71.1 59.9 21.8
## 222 62.3 29.7 84.3 53.9 15.4
## 223 63.2 9.3 84.4 53.9 21.5
## 224 46.6 39.6 84.2 56.4 21.0
## 225 67.8 34.5 78.5 60.0 36.9
## 226 47.6 19.1 60.0 64.9 29.4
## 227 67.9 14.2 62.9 72.6 22.4
## 228 62.9 20.3 84.0 58.9 19.2
## 229 68.6 34.8 59.3 54.7 48.2
## 230 48.7 18.2 81.3 84.4 27.2
## 231 61.1 2.5 68.6 76.9 15.0
## 232 54.4 23.3 70.2 73.4 24.8
## 233 42.4 34.3 65.5 85.4 31.1
## 234 61.8 11.1 72.2 53.3 40.0
## 235 81.3 22.1 85.5 89.6 29.0
## 236 52.2 10.3 80.4 79.7 31.3
## 237 57.6 26.4 72.2 68.7 17.5
## 238 68.3 23.4 85.4 72.5 39.6
## 239 60.4 24.6 94.9 65.7 31.2
## 240 83.3 17.5 74.9 63.7 38.8
## 241 68.8 16.5 80.6 46.1 48.6
## 242 87.9 35.5 78.3 73.9 32.1
## 243 64.2 26.3 63.3 90.7 26.6
## 244 46.4 22.8 63.9 90.1 20.8
## 245 65.1 0.0 98.0 67.0 23.1
## 246 48.9 9.7 85.6 63.7 19.0
## 247 47.9 34.2 83.9 67.2 34.4
## 248 57.7 20.1 76.8 78.8 12.4
## 249 71.2 10.5 80.8 82.7 9.1
## 250 64.1 19.7 76.8 69.9 22.0
## 251 69.8 20.5 75.5 78.4 43.0
## 252 71.0 19.0 77.7 64.7 31.5
## 253 57.9 31.0 62.9 71.0 24.1
## 254 31.2 43.3 100.0 79.2 38.0
## 255 69.6 36.9 84.6 100.0 25.8
## 256 86.0 40.0 72.0 64.4 28.6
## 257 60.7 21.0 89.6 57.0 30.9
## 258 43.7 35.1 67.5 70.3 43.1
## 259 55.6 13.2 95.5 60.1 45.5
## 260 48.8 23.3 70.3 72.6 45.4
## 261 59.5 51.2 47.1 69.1 41.7
## 262 68.1 24.8 91.4 80.8 34.0
## 263 70.4 21.3 88.6 53.5 52.6
## 264 62.8 18.0 65.7 50.5 54.5
## 265 48.8 34.3 58.1 83.8 45.9
## 266 69.8 50.3 53.0 67.7 27.4
## 267 76.2 29.9 71.8 53.9 17.5
## 268 87.0 21.9 53.0 52.1 32.3
## 269 54.1 49.9 65.2 58.4 23.3
## 270 65.7 15.7 77.0 72.1 33.3
## 271 74.3 38.6 85.3 58.7 23.4
## 272 58.6 19.9 65.2 81.1 41.1
## 273 66.3 12.7 73.5 67.9 56.3
## 274 62.6 39.6 78.1 67.4 41.6
## 275 58.4 3.4 95.6 64.8 38.9
## 276 62.0 21.3 63.5 61.4 30.4
## 277 71.6 25.2 55.7 63.0 56.2
## 278 64.9 19.7 52.9 78.9 29.0
## 279 54.4 5.3 81.7 48.9 46.5
## 280 33.1 17.3 74.3 78.0 48.9
## 281 50.5 6.3 84.3 60.8 31.9
## 282 59.8 48.1 73.3 73.8 0.0
## 283 29.8 2.7 79.7 64.3 26.0
## 284 86.5 0.0 77.7 48.9 35.5
## 285 42.1 33.4 75.3 69.2 14.4
## 286 46.1 35.9 67.5 57.3 14.5
## 287 77.5 22.6 90.1 78.5 40.9
## 288 33.7 22.5 70.1 55.1 0.0
## 289 68.9 33.7 83.0 43.5 32.8
## 290 55.9 41.8 77.0 54.8 26.8
## 291 36.8 51.3 51.1 54.9 52.5#install.packages("factoextra")
library(factoextra)## Welcome! Want to learn more? See two factoextra-related books at https://goo.gl/ve3WBadf <- customer_movie_rating
df <- na.omit(df)
df <- scale(df)
head(df)## Horror Romcom Action Comedy Fantasy
## 1 0.5716500 -0.0791958 -0.01515773 -0.6986964 0.5587483
## 2 0.9696365 0.8275698 0.48122544 -1.7277543 1.0042660
## 3 0.4690762 0.7510247 -0.23507432 -0.1245439 -0.3550175
## 4 1.6630357 -0.6032357 0.56919208 -1.3744297 -0.2413650
## 5 1.0434897 -1.8397343 -0.04657438 -0.2967896 -0.2550033
## 6 0.4772821 1.3987144 1.78187502 -0.3718711 -0.2322728distance <- get_dist(df)
?get_dist## starting httpd help server ... donefviz_dist(distance)
k2 <- kmeans(df, centers = 3, nstart = 25)
str(k2)## List of 9
## $ cluster : Named int [1:291] 3 3 1 3 3 3 3 3 3 3 ...
## ..- attr(*, "names")= chr [1:291] "1" "2" "3" "4" ...
## $ centers : num [1:3, 1:5] 0.0441 -1.3916 0.9064 -0.4249 1.2101 ...
## ..- attr(*, "dimnames")=List of 2
## .. ..$ : chr [1:3] "1" "2" "3"
## .. ..$ : chr [1:5] "Horror" "Romcom" "Action" "Comedy" ...
## $ totss : num 1450
## $ withinss : num [1:3] 206 156 191
## $ tot.withinss: num 554
## $ betweenss : num 896
## $ size : int [1:3] 114 72 105
## $ iter : int 3
## $ ifault : int 0
## - attr(*, "class")= chr "kmeans"?kmeanslibrary(factoextra)
fviz_cluster(k2, data = df)
df %>%
as_tibble() %>%
mutate(cluster = k2$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Horror, Romcom, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k2$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Horror, Action, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k2$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Horror, Comedy, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k2$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Horror, Fantasy, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k2$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Romcom, Action, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k2$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Romcom, Comedy, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k2$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Romcom, Fantasy, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k2$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Action, Comedy, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k2$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Action, Fantasy, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k2$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Comedy, Fantasy, color = factor(cluster), label = state)) +
geom_text()
k3 <- kmeans(df, centers = 3, nstart = 25)
k4 <- kmeans(df, centers = 4, nstart = 25)
k5 <- kmeans(df, centers = 5, nstart = 25)
# plots to compare
p1 <- fviz_cluster(k2, geom = "point", data = df) + ggtitle("k = 2")
p2 <- fviz_cluster(k3, geom = "point", data = df) + ggtitle("k = 3")
p3 <- fviz_cluster(k4, geom = "point", data = df) + ggtitle("k = 4")
p4 <- fviz_cluster(k5, geom = "point", data = df) + ggtitle("k = 5")
library(gridExtra)
grid.arrange(p1, p2, p3, p4, nrow = 2)
df %>%
as_tibble() %>%
mutate(cluster = k3$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Horror, Romcom, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k3$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Horror, Action, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k3$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Horror, Comedy, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k3$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Horror, Fantasy, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k3$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Romcom, Action, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k3$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Romcom, Comedy, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k3$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Romcom, Fantasy, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k3$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Action, Comedy, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k3$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Action, Fantasy, color = factor(cluster), label = state)) +
geom_text()
df %>%
as_tibble() %>%
mutate(cluster = k3$cluster,
state = row.names(customer_movie_rating)) %>%
ggplot(aes(Comedy, Fantasy, color = factor(cluster), label = state)) +
geom_text()
## Ideal Cluster Numbers
set.seed(123)
# function to compute total within-cluster sum of square
wss <- function(k) {
kmeans(df, k, nstart = 10 )$tot.withinss}
# Compute and plot wss for k = 1 to k = 15
k.values <- 1:15
# extract wss for 2-15 clusters
wss_values <- map_dbl(k.values, wss)
plot(k.values, wss_values,
type="b", pch = 19, frame = FALSE,
xlab="Number of clusters K",
ylab="Total within-clusters sum of squares")
set.seed(123)
fviz_nbclust(df, kmeans, method = "wss")
#Average Silhouette Method
fviz_nbclust(df, kmeans, method = "silhouette")
#gap statistic
library(cluster)
# compute gap statistic
set.seed(123)
gap_stat <- clusGap(df, FUN = kmeans, nstart = 25,
K.max = 10, B = 50)
# Print the result
print(gap_stat, method = "firstmax")## Clustering Gap statistic ["clusGap"] from call:
## clusGap(x = df, FUNcluster = kmeans, K.max = 10, B = 50, nstart = 25)
## B=50 simulated reference sets, k = 1..10; spaceH0="scaledPCA"
## --> Number of clusters (method 'firstmax'): 3
## logW E.logW gap SE.sim
## [1,] 5.354803 5.723527 0.3687242 0.01293288
## [2,] 5.058695 5.541902 0.4832072 0.01242534
## [3,] 4.899418 5.461444 0.5620263 0.01243341
## [4,] 4.853367 5.396459 0.5430921 0.01258290
## [5,] 4.809464 5.346107 0.5366437 0.01169743
## [6,] 4.775626 5.302812 0.5271867 0.01142782
## [7,] 4.741856 5.264109 0.5222532 0.01158757
## [8,] 4.708943 5.228317 0.5193740 0.01165380
## [9,] 4.676318 5.196718 0.5204000 0.01188042
## [10,] 4.649985 5.167409 0.5174246 0.01257006fviz_gap_stat(gap_stat)
set.seed(123)
final <- kmeans(df, 3, nstart = 25)
print(final)## K-means clustering with 3 clusters of sizes 114, 72, 105
##
## Cluster means:
## Horror Romcom Action Comedy Fantasy
## 1 0.04409655 -0.4249402 0.3456929 0.5053192 -0.78478335
## 2 -1.39161333 1.2100503 -1.1767013 0.8158762 1.29578460
## 3 0.90637289 -0.3683851 0.4315572 -1.1080902 -0.03648752
##
## Clustering vector:
## 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
## 3 3 1 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3
## 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
## 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3
## 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
## 3 3 3 3 3 3 3 3 3 1 3 3 3 3 3 3 3 3 3 3
## 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80
## 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3
## 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
## 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3
## 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120
## 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
## 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140
## 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
## 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160
## 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
## 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180
## 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1
## 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200
## 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
## 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220
## 1 1 1 1 3 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
## 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240
## 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
## 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260
## 3 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
## 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280
## 2 1 3 3 1 1 1 3 1 1 1 1 1 1 1 1 1 1 1 1
## 281 282 283 284 285 286 287 288 289 290 291
## 1 1 1 3 1 1 1 1 3 1 2
##
## Within cluster sum of squares by cluster:
## [1] 205.8360 156.4432 191.4703
## (between_SS / total_SS = 61.8 %)
##
## Available components:
##
## [1] "cluster" "centers" "totss" "withinss" "tot.withinss"
## [6] "betweenss" "size" "iter" "ifault"customer_movie_rating %>%
mutate(Cluster = final$cluster) %>%
group_by(Cluster) %>%
summarise_all("mean")## # A tibble: 3 x 6
## Cluster Horror Romcom Action Comedy Fantasy
## <int> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 1 59.6 24.0 74.3 68.0 28.3
## 2 2 24.6 51.8 50.1 75.0 74.1
## 3 3 80.7 25.0 75.7 31.4 44.8#Hierarchical
library(tidyverse) # data manipulation
library(cluster) # clustering algorithms
library(factoextra) # clustering visualization
library(dendextend) # for comparing two dendrograms##
## ---------------------
## Welcome to dendextend version 1.15.1
## Type citation('dendextend') for how to cite the package.
##
## Type browseVignettes(package = 'dendextend') for the package vignette.
## The github page is: https://github.com/talgalili/dendextend/
##
## Suggestions and bug-reports can be submitted at: https://github.com/talgalili/dendextend/issues
## Or contact: <tal.galili@gmail.com>
##
## To suppress this message use: suppressPackageStartupMessages(library(dendextend))
## ---------------------##
## Attaching package: 'dendextend'## The following object is masked from 'package:stats':
##
## cutree# Dissimilarity matrix
d <- dist(df, method = "euclidean")
# Hierarchical clustering using Complete Linkage
hc1 <- hclust(d, method = "complete" )
# Plot the obtained dendrogram
plot(hc1, cex = 0.6, hang = -1)
# Compute with agnes
hc2 <- agnes(df, method = "complete")
# Agglomerative coefficient
hc2$ac## [1] 0.8962888# methods to assess
m <- c( "average", "single", "complete", "ward")
names(m) <- c( "average", "single", "complete", "ward")
# function to compute coefficient
ac <- function(x) {
agnes(df, method = x)$ac
}
map_dbl(m, ac)## average single complete ward
## 0.8210505 0.5923947 0.8962888 0.9794408hc3 <- agnes(df, method = "ward")
pltree(hc3, cex = 0.6, hang = -1, main = "Dendrogram of agnes") 
## DIANA
# compute divisive hierarchical clustering
hc4 <- diana(df)
# Divise coefficient; amount of clustering structure found
hc4$dc## [1] 0.8831184## [1] 0.8514345
# plot dendrogram
pltree(hc4, cex = 0.6, hang = -1, main = "Dendrogram of diana")
working with the result
# Ward's method
hc5 <- hclust(d, method = "ward.D2" )
# Cut tree into 4 groups
sub_grp <- cutree(hc5, k = 4)
sub_grp <- cutree(hc5, k = 6)
?cutree
# Number of members in each cluster
table(sub_grp)## sub_grp
## 1 2 3 4 5 6
## 43 38 63 67 49 31customer_movie_rating %>%
mutate(cluster = sub_grp) %>%
head## Horror Romcom Action Comedy Fantasy cluster
## 1 72.5 29.9 68.6 40.7 57.9 1
## 2 82.2 45.3 76.5 17.4 67.7 1
## 3 70.0 44.0 65.1 53.7 37.8 2
## 4 99.1 21.0 77.9 25.4 40.3 3
## 5 84.0 0.0 68.1 49.8 40.0 3
## 6 70.2 55.0 97.2 48.1 40.5 2plot(hc5, cex = 0.6)
rect.hclust(hc5, k = 6, border = 2:5)
# Hierarchical Diana result
d2<- customer_movie_rating %>%
mutate(cluster = sub_grp) %>%
head
customer_movie_rating %>%
mutate(cluster = sub_grp) %>%
group_by(cluster) %>%
summarise_all("mean")## # A tibble: 6 x 6
## cluster Horror Romcom Action Comedy Fantasy
## <int> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 1 81.3 26.5 82.6 30.3 54.4
## 2 2 50.9 39.3 72.2 64.0 32.7
## 3 3 81.4 24.3 69.0 33.7 36.3
## 4 4 23.7 52.0 50.2 75.8 75.9
## 5 5 61.9 18.2 83.2 67.8 30.3
## 6 6 60.3 17.9 63.0 70.6 26.9#K means result
customer_movie_rating %>%
mutate(Cluster = final$cluster) %>%
group_by(Cluster) %>%
summarise_all("mean") ## # A tibble: 3 x 6
## Cluster Horror Romcom Action Comedy Fantasy
## <int> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 1 59.6 24.0 74.3 68.0 28.3
## 2 2 24.6 51.8 50.1 75.0 74.1
## 3 3 80.7 25.0 75.7 31.4 44.8Berdasarkan clustering yang dilakukan di atas, dapat disimpulkan:
Optimal jumlah cluster yang optimal adalah sebesar 3 cluster atau K-3
Untuk Cluster 1, rating movie tertinggi terdapat pada genre Action dan yang terendah adalah genre Romcom.
Untuk Cluster 2, rating movie tertinggi terdapat pada genre Comedy dan yang terendah adalah genre Horror.
Untuk Cluster 3, rating movie tertinggi terdapat pada genre Horror dan yang terendah adalah genre Romcom.
Pada semua cluster yang ada, Genre Horror memperoleh rating paling tinggi sedangkan genre Romcom memperoleh rating paling rendah.